In many imaging system, it is desirable to optimize exposure and brightness. However, the captured image is often not pleasing, because of exposure and brightness.
A common problem that occurs in many digital imaging systems is that the images are too dark or too light. Often these problems result from errors introduced by the exposure control system in a digital image capture device. Many methods have been developed to improve the brightness of digital images. Such techniques are often referred to as “scene balance algorithms.” Scene balance algorithms can vary widely in their complexity, as well as in the accuracy of their results. They typically involve analyzing the distribution of overall exposure levels and relative color signal levels in an image to determine the appropriate level of brightness and color balance compensation that is needed.
Examples of prior art brightness adjustment methods are described in U.S. Pat. Nos. 4,101,217, 4,707,119, 4,984,013; 4,945,406, 5,016,043, 6,636,646, 6,845,181, 7,129,980 and 7,289,154.
Conventional scene balance algorithms apply global brightness and color balance adjustments to an entire image. Many times only portions of an image may suffer from being too dark or too light. For example, an object in the shadows, or an object that is in the background of a flash image may be too dark, or an image in the foreground of a flash image may be too light. U.S. Pat. Nos. 6,275,605 and 7,043,090 describe methods for adjusting the tone scale of an image so that different brightness adjustments can be applied to different portions of an image. The tone scale adjustments are based on an analysis of the two-dimensional digital images. Sometimes these algorithms suffer from artifacts where the amount of tone scale adjustment is not constant within an object.
Most prior art brightness adjustment methods determine brightness correction values based on an analysis of conventional two-dimensional digital images. US Patent Application Publication No. 2007/0126921 describes an exposure and tone scale adjustment method that uses range information for an image. With this method, the range information is used to determine a weighting function that places additional importance on image regions that are likely to contain important objects. The weighting function is then used to determine an overall exposure adjustment amount for the image. While this method makes it more likely that the main subject of an image is properly exposed, it does not address the problem that different objects in an image may need different exposure adjustments.